Sentiment and intent analysis for customizing suggestions using user-specific information

Abstract

Systems and processes for operating an intelligent automated assistant to provide customized suggestions based on user-specific information are provided. An example method includes obtaining impressions and performing, based on the impressions, at least one of: analyzing sentiment of at least a portion of the impressions; and predicting user intent based on at least a portion of the impressions. The method further includes determining a plurality of concepts based on the obtained impressions; and weighing the plurality of concepts based on context associated with obtaining the impressions and based on at least one of a sentiment analysis result or a predicted user intent. The method further includes generating, based on the one or more weighted concepts, a representation of a collection of user-specific information; and facilitating to provide one or more suggestions to the user based on the representation of the collection of user-specific information.

What is claimed is: 
     
         1 . A method for facilitating to provide one or more suggestions to a user, comprising:
 at an electronic device with one or more processors and memory:   obtaining impressions associated with at least one of the electronic device or additional electronic devices communicatively coupled to the electronic device;   performing, based on the impressions, at least one of:
 analyzing sentiment of at least a portion of the impressions; and 
 predicting user intent based on at least a portion of the impressions; 
   determining a plurality of concepts based on the obtained impressions;   weighing the plurality of concepts based on context associated with obtaining the impressions and based on at least one of a sentiment analysis result or a predicted user intent;   generating, based on the plurality of weighted concepts, a representation of a collection of user-specific information; and   facilitating to provide one or more suggestions to the user based on the representation of the collection of user-specific information.   
     
     
         2 . The method of  claim 1 , wherein obtaining the impressions comprises:
 collecting data items from one or more data sources associated with at least one of the electronic device or the additional electronic devices communicatively coupled to the electronic device;   determining whether the collected data items represent one or more inputs from the user; and   in accordance with a determination that the collected data items represent one or more inputs from the user, including the collected data items in the impressions.   
     
     
         3 . The method of  claim 2 , wherein the collected data items include at least one of: one or more files, one or more search queries, and one or more user inputs. 
     
     
         4 . The method of  claim 1 , wherein analyzing sentiment of at least a portion of the impressions comprises:
 generating tokens based on one or more data items represented by the at least a portion of the impressions, wherein the one or more data items comprising natural language text;   processing the tokens using a first machine learning model pre-trained for identifying sentiment; and   predicting sentiment of the one or more data items based on results of processing the tokens.   
     
     
         5 . The method of  claim 4 , wherein the one or more data items include a plurality of data items, further comprising, grouping the plurality of data items before generating the tokens. 
     
     
         6 . The method of  claim 1 , wherein predicting the user intent based on at least a portion of the impressions comprises:
 generating a plurality of tokens based on one or more data items represented by the at least a portion of the impressions, wherein the one or more data items comprising natural language text;   processing the plurality of tokens using a second machine learning model pre-trained for user intent prediction; and   predicting the user intent based on results of processing the tokens.   
     
     
         7 . The method of  claim 6 , wherein the one or more data items include a plurality of data items, further comprising: grouping the plurality of data items before generating the tokens. 
     
     
         8 . The method of  claim 6 , wherein processing the plurality of tokens using the second machine learning model pre-trained for user intent prediction comprises:
 determining one or more polarities associated with the plurality of tokens;   determining one or more probabilities associated with the one or more polarities; and   determining whether event information is present based on the probabilities associated with the polarities.   
     
     
         9 . The method of  claim 8 , wherein processing the plurality of tokens using the second machine learning model pre-trained for user intent prediction further comprises:
 in accordance with a determination that event information is present, determining an event location or an entity associated with the event location.   
     
     
         10 . The method of  claim 8 , wherein processing the plurality of tokens using the second machine learning model pre-trained for user intent prediction further comprises:
 in accordance with a determination that event information is present, determining an event time; and   determining whether the event time indicates a past event or a future event.   
     
     
         11 . The method of  claim 8 , wherein predicting the user intent based on results of processing the tokens comprises:
 comparing the one or more probabilities associated with the one or more polarities with at least one probability threshold; and   predicting the user intent based on a result of comparing the one or more probabilities associated with the one or more polarities with at least one probability threshold.   
     
     
         12 . The method of  claim 1 , wherein determining the plurality of concepts based on the impressions comprises determining at least one of:
 one or more topics;   one or more entities;   a user identity; and   one or more recurrent user inputs.   
     
     
         13 . The method of  claim 1 , further comprising, prior to weighing the plurality of concepts, assigning a score to each of the plurality of concepts, the score representing a likelihood the concept is to be used in providing suggestions to the user. 
     
     
         14 . The method of  claim 1 , wherein weighing the plurality of concepts comprises, for at least one concept of the plurality of concepts:
 determining, based on a result of the sentiment analysis, whether sentiment of at least a portion of the impressions is positive; and   in accordance with a determination that sentiment associated with the at least one concept is positive, increasing one or more scores assigned to the at least one concept.   
     
     
         15 . The method of  claim 1 , wherein weighing the plurality of concepts comprises, for at least one concept of the plurality of concepts:
 determining whether the predicted user intent corresponds to an acceptance polarity; and   in accordance with a determination that the predicted user intent corresponds to an acceptance polarity, increasing one or more scores associated with the at least one concept.   
     
     
         16 . The method of  claim 1 , wherein weighing the plurality of concepts comprises, for at least one concept of the plurality of concepts:
 adjusting one or more scores of the at least one concept based on at least one of a timing or a location associated with obtaining at least a portion of the impressions.   
     
     
         17 . The method of  claim 16 , wherein adjusting one or more scores of the at least one concept based on at least one of a timing or a location associated with obtaining the impressions comprises:
 determining whether the timing indicates that user activities are persistent over a pre-defined time period, wherein the impressions are obtained based on the user activities;   in accordance with a determination that the timing indicates that the user activities are persistent over a pre-defined time period, increasing one or more scores of the at least one concept.   
     
     
         18 . The method of  claim 1 , wherein generating, based on the one or more weighted concepts, the representation of the collection of user-specific information comprises:
 performing at least one of a categorizing or a ranking the one or more weighted concepts; and   generating the representation of the collection of user-specific information based on results of performing at least one of categorizing or ranking of the one or more weighted concepts.   
     
     
         19 . The method of  claim 18 , wherein performing at least one of a categorizing or a ranking the one or more weighted concepts comprises:
 grouping two or more of the weighted concepts, wherein the grouped two or more concepts are used for providing suggestions to the user.   
     
     
         20 . The method of  claim 19 , wherein grouping of the two or more of the weighted concepts is based on at least one of a timing or a source of at least one obtained impressions from which the concepts are determined. 
     
     
         21 . The method of  claim 19 , wherein grouping of the two or more of the weighted concepts is based on a classification of the two or more of the weighted concepts. 
     
     
         22 . The method of  claim 1 , wherein facilitating to provide one or more suggestions to the user based on the representation of the collection of user-specific information comprises:
 receiving, from a querying client associated with at least one of the electronic device or the additional electronic devices communicatively coupled to the electronic device, one or more queries of user-specific information;   in response to the one or more queries, determining the user-specific information based on the representation of the collection of user-specific information; and   providing the user-specific information to the querying client.   
     
     
         23 . A non-transitory computer-readable storage medium storing one or more programs for providing word correction, the one or more programs comprising instruction, which when executed by one or more processors of an electronic device, cause the electronic device to:
 obtain impressions associated with at least one of the electronic device or additional electronic devices communicatively coupled to the electronic device;   perform, based on the impressions, at least one of:
 analyzing sentiment of at least a portion of the impressions; and 
 predicting user intent based on at least a portion of the impressions; 
   determine a plurality of concepts based on the obtained impressions;   weigh the plurality of concepts based on context associated with obtaining the impressions and based on at least one of a sentiment analysis result or a predicted user intent;   generate, based on the plurality of weighted concepts, a representation of a collection of user-specific information; and   facilitate to provide one or more suggestions to the user based on the representation of the collection of user-specific information.   
     
     
         24 . An electronic device, comprising:
 one or more processors;   memory; and   one or more programs stored in memory, the one or more programs including instructions for:   obtaining impressions associated with at least one of the electronic device or additional electronic devices communicatively coupled to the electronic device,   performing, based on the impressions, at least one of:
 analyzing sentiment of at least a portion of the impressions; and 
 predicting user intent based on at least a portion of the impressions; 
   determining a plurality of concepts based on the obtained impressions;   weighing the plurality of concepts based on context associated with obtaining the impressions and based on at least one of a sentiment analysis result or a predicted user intent;   generating, based on the plurality of weighted concepts, a representation of a collection of user-specific information; and   facilitating to provide one or more suggestions to the user based on the representation of the collection of user-specific information.   
     
     
         25 . An electronic device, comprising:
 means for obtaining impressions associated with at least one of the electronic device or additional electronic devices communicatively coupled to the electronic device;   means for performing, based on the impressions, at least one of:
 analyzing sentiment of at least a portion of the impressions; and 
 predicting user intent based on at least a portion of the impressions; 
   means for determining a plurality of concepts based on the obtained impressions;   means for weighing the plurality of concepts based on context associated with obtaining the impressions and based on at least one of a sentiment analysis result or a predicted user intent;   means for generating, based on the plurality of weighted concepts, a representation of a collection of user-specific information; and   means for facilitating to provide one or more suggestions to the user based on the representation of the collection of user-specific information.